Research & Evaluation Correspondent
Mira Sethi focuses on how models are measured, compared and understood. Her coverage examines evaluation design, reproducibility and the assumptions behind claims of progress. She looks for the caveat that changes the headline.
711 published articles · Showing 141–160
A groundbreaking new benchmark, IRC-Bench, has just been unveiled, designed to challenge artificial intelligence systems to interpret the subtle, implicit references within personal memories arXiv CS. AI....
Read article →Two new research papers emerging from arXiv today are set to reshape how founders approach building with Large Language Model (LLM) agents, directly addressing critical bottlenecks in both coding efficiency and advanced reasoning capabilities. This isn't just academic chatter; it...
Read article →Boox, a company known for its e-readers, has launched the Tappy, a new multi-functional remote control designed to enhance the reading and content consumption experience. This tiny device, priced at $25....
Read article →Anthropic, the AI powerhouse co-founded by former OpenAI research leaders, has shattered expectations by revealing a staggering $30 billion revenue run rate, fueled by an 'unbelievable' 80x growth surge. This isn't just a number; it's a defiant roar in the arena of deep tech, a t...
Read article →The relentless demand for tools that empower the next generation of digital builders is undeniable, and two key developments this week underscore the market's insatiable appetite. Asus has unveiled a specialized secondary touchscreen display, the ROG Strix XG129C, aimed directly ...
Read article →Cutting-edge research from arXiv reveals critical vulnerabilities in the rapidly evolving landscape of artificial intelligence, exposing new attack vectors against multimodal systems and introducing sophisticated frameworks for evaluating AI defenses. These two papers, published ...
Read article →The National Highway Traffic Safety Administration (NHTSA) has launched a formal investigation into Avride, an Uber partner, following more than a dozen crashes and one minor injury linked to its self-driving vehicles, intensifying scrutiny on the autonomous vehicle sector. Simul...
Read article →A pair of groundbreaking research papers, published today on arXiv CS. LG, reveal significant advancements in tackling the inherent inefficiencies of Mixture-of-Experts (MoE) architectures, a critical component in scaling large language models....
Read article →New research emerging from arXiv this week is poised to fundamentally shift how startups grapple with uncertainty, offering advanced machine learning paradigms designed for the relentless, dynamic environments they inhabit. Two distinct papers, published on May 8, 2026, delve int...
Read article →A flurry of new research papers on arXiv, all published today, May 8, 2026, signals a critical leap forward in AI interpretability and explainability—a domain vital for any founder pushing to deploy robust, trustworthy models into the real world. These advancements tackle everyth...
Read article →Two new research papers, published today on arXiv, are set to fundamentally shift the landscape for large language model (LLM) development, addressing critical bottlenecks in scalability and computational efficiency that have long challenged builders. These papers, one on advance...
Read article →A trio of new research papers published today on arXiv CS. LG signals a significant leap in foundational AI capabilities, addressing critical challenges in multi-agent decision-making, efficient reinforcement learning, and sophisticated multi-stage planning....
Read article →A new research paper published on arXiv today introduces a novel AI algorithm poised to revolutionize the design and optimization of fluorescent proteins. This breakthrough, detailed in arXiv:2605....
Read article →Today's torrent of research from arXiv reveals a pivotal moment for Large Language Models (LLMs): a dual focus on shoring up their foundational reliability and dramatically boosting their efficiency. Fresh papers published today highlight critical progress in making LLMs more tru...
Read article →Today's AI funding landscape revealed a stark divergence, with established player Kodiak AI closing a challenging $100 million round at a steep discount, sending its stock plunging 37%, even as a new contender, Pit AI, co-founded by the visionaries behind European scooter giant V...
Read article →In a week marked by strategic recalibrations from tech's heavyweights, Xbox CEO Asha Sharma rallied hundreds of employees with a 'return of Xbox' mandate, signaling a renewed focus after two tumultuous years. Simultaneously, Disney Plus CEO Josh D'Amaro unveiled an ambitious visi...
Read article →Ethan Shaotran, a figure well-known for his work at DOGE, has launched a new defense technology startup named Blitz Industries. This significant pivot by an established tech leader signals a growing trend of seasoned talent entering critical sectors, driven by the dual promise of...
Read article →A torrent of new research published today on arXiv CS. AI signals a critical maturation point for AI agents, pushing the field beyond theoretical demonstrations toward robust, safe, and truly proactive real-world deployments....
Read article →The U.S. Food and Drug Administration (FDA) has reportedly suppressed studies confirming the benefits of critical vaccines, including COVID-19 and Shingrix, under the Trump administration. This alarming development, aligning with anti-vacci...
Read article →SpaceX, already a titan in rocketry and satellite internet, is now poised to make an unprecedented leap into high-stakes semiconductor manufacturing, considering an initial investment of $55 billion in a Texas chip factory that could ultimately reach $119 billion TechCrunch. This...
Read article →